Friday, 25 April 2014

Close to Home: the problem of familiar surroundings

Another post based on an old YSDC post.

Although my examples are specifically about a school scenario, and based on Call of Cthulhu, I think they're generally applicable to any situation where the players are familiar with their surroundings and the NPCs. It might be a house, a small village, a ship or a military base. These situations seem to threaten a lot of problems for the GM in maintaining the mystery, in keeping investigative control with the players rather than doling out instructions a piece at a time, and in trying to maintain some realism.

At the time, I was working on a now-dormant Call of Cthulhu scenario set in a school. This quickly ran into the point that the characters are intimately familiar with the location and its inhabitants. Of course, there are some areas they don't visit, and any character can have secrets, but player characters and NPCs alike should realistically have a very good knowledge of things around them.

There must be existing scenarios where the action happens somewhere very familiar to the investigators, but I don't know them. Certainly most scenarios assume you aren't previously familiar with plot-relevant locations, but get called to visit them and discover their secrets. In many cases you move between several locations, so even if hooks appear somewhere you know, you'll be led elsewhere. Although investigators may know a certain city, or even a district, they don't tend to know the houses, graveyards, docks or occult bookshops where the plot unfolds. Information about these places is gradually discovered over the course of the scenario, almost always starting from zero.

There are three main consequences of a location being new to the PCs.

Firstly, once they find out it "exists" within the scenario, investigators are very likely to go there. There is a natural progression from total ignorance of the "geography" of the scenario, where the party find out about the places there are to visit and check them out. If you find a clue even hinting at a connection to another location that you've never heard of, you're very likely to go there. This is largely the tacit understanding that if information uncovered in a game (even an investigative game) isn't relevant, it wouldn't be there. It's not always true, or sometimes players and GMs ascribe different importance to a throwaway fact, but there's limited cost to going, so typically you might as well.

Secondly, information can be given out naturally without beating anyone with the Hint stick. Because the investigators are new to the place, they can be assumed to be paying some amount of attention. They don't know what a place is like, so everything they might see will get some attention even if they aren't actively looking for it. What kind of buildings are there? How many people? Is it in good repair? Details that you'd skim over in a familiar place need mentioning, and these may give clues or hint at dangers. You can also include things that seem strange or scary but habitual visitors would get used to, like odd architectural features, strange noises or unusual inhabitants.

Thirdly, characters will actively investigate it, either by research or in person. What's its history? Any new stories, rumours or legends? What does it look like? Who is there? Are there any interesting features? Any treasure? What kind of atmosphere does the place have? These questions are likely to come up naturally, and if the answers are interesting, it's reasonable that they go and investigate more closely.

Familiar locations

I think that a scenario set somewhere the investigators are already intimately familiar with doesn't work quite the same.

For a start, they should already know all the major locations that exist, so it doesn't make sense for them to discover them sequentially during the plot. They start off with a map or something and some blurb. This also means there is no particular incentive for them to go somewhere just because it is mentioned; their characters already know what that place is like, so it makes no sense in character to go and explore it. Lacking that sense of novelty, there's also less out-of-character motivation to do so.

Secondly, giving out meaningful information in a useful way becomes more difficult. Because the investigators aren't exploring the area from scratch, they don't get a little bit of information at once; they basically have to start off with a big pile of information about the area, of which some is plot-relevant and some isn't. Unless you include a huge amount of unnecessary detail (which is hard work and may be annoying), it will be fairly obvious which information is relevant and which isn't, so they will be looking out for links right from the start. This isn't necessarily a problem, but it depends what sort of game you're playing and what the group's tastes are. Many players don't like wading through large amounts of information, even things like general setting and history, so asking them to read up on a scenario location isn't likely to go down well.

If you don't give out an initial infodump, you have a different problem. The investigators are effectively rendered amnesiac, with little idea of what the world around them is like. They may end up wasting a lot of player time traipsing round looking at everything, and finding out about the local culture, rules, residents and so on. They don't know, until they ask, that there was a fire in the chemistry lab last term, which is an important clue to the plot. They don't know of the local tensions between the Baron and the Guild of Miners, which might be behind the crime wave. They have no ambient knowledge of the NPCs, either in terms of their social roles (is there a blacksmith? a forensic accountant? a Martian?) or their personalities. This means that whenever they find any information, you have to basically tell them the clue-value of that information.

Basically, either option means it's quite difficult to do subtlety.

Thirdly, it's less natural to actively investigate things. Why would you go prying around the dormitory you sleep in every night, or exploring the chapel you've been visiting for years? Why would you wonder what the inscription is under the paintings when you pass them every day on the way to lessons? If the Headmaster mentions the name of an old boy in his morning assembly, you've probably heard it at least once a term (few schools have many significant old boys) so why would you be curious about him? Because things are not new, they aren't intriguing. This is a real-life issue: I very rarely go to any tourist attractions where I live: I can go there any time, so I don't go at all, not even once. Since the PC don't want to investigate, it's hard for the players to justify it plausibly, and hard for the GM to provide information in small doses.

Another difficulty with this fixed, known setting is that very little can be hidden, and especially people. In an ordinary scenario, once a suspicious name crops up, that's not enough: you need to track down the person too. They might be hard to find, they might use a pseudonym, or they might run away if they suspect you're coming. However, when you're talking about a fellow pupil at the school, things are a bit different. If your suspicion lights on someone, it's fairly easy to find them, at which point things typically come down to deception or fighting. They can't really disappear without becoming incredibly suspicious, nor can they assume a false identity. You can easily find out where they sleep (or where a teacher's office is) and track their movements. Finding and searching their possessions is relatively simple (though not necessarily easy). This limits the investigative aspects of the game.

Dissonance of Salience

I think these are symptoms of a more general problem I'm going to call Dissonance of Salience (cf. cognitive dissonance). Information that is very salient to the player is not salient in-character, and so it becomes difficult to convey it in a way that's natural and doesn't undermine the game. This problem is most prevalent in terms of investigation, but can also crop up in terms of cultural and behavioural norms: what is a normal interaction between a dwarven cleric and an elven wizard, and how should any discrepancies be interpreted? Baring your teeth is aggressive to Plutonian lizard-people, and the Space Corps know that, but do you best convey to the players that smiling is bad in a way that leads to a better game experience, rather than effectively rendering the point moot ("okay, we don't smile then")? You don't want to dictate to players how they play their characters, but having interesting and consistent patterns of interaction and social relationships help make a setting feel alive. If dwarves and orks are supposed to hate each other, or people treat robots like treasured servants rather than slaves, then you want players to be aware of that, so any departures on their part are intentional and meaningful.

Things like social mores and superstitions are another good example. Major departures by NPCs would obviously attract PC attention (running around naked during a funeral, say), but more subtle breaches that might hint at oddness are harder to convey without beating anyone over the head with the Clue Bat. Again, if you want to aim for verisimilitude, then these kinds of details are important - especially in things like historical settings where there's an objective reality to compare against. A vagrant and a common soldier are not getting a private meeting with the Duchess of Avon and no servants present; they won't even get to leave her a note.

An example

So to stick with my school example: let's say there was a fire in the chemistry lab last term, something that would be known to the whole school.

If you mention this at the start, so that investigators have the appropriate background knowledge, it's obvious that this is linked to the scenario. If you don't mention it until players ask "was there a fire anywhere last term?", you're cheating them of information they would actually have, which makes it unlikely they'll deduce anything useful. If you give them this information when they find a clue, it's a massive nudge-wink; essentially you're handing out a chunk of plot, rather than a clue. This is a Sherlock Holmes technique, where there's no way for them to interpret any clues because you hold back crucial information until you want them to know the answer.

So let's say the investigators find a labcoat lying around, and examine it. "You find a box of matches." Two crucial bits of background knowledge apply: there was a fire in the chemistry lab last term, and pupils are not allowed matches. There isn't necessarily a connection between the two (maybe it's just an illicit smoker, a red herring) but it's a big clue. There are several ways to handle this:

  • You don't mention this background information at all, unless someone asks you. Result: it's very unlikely anyone works this out. Matches are fairly normal items, not intrinsically suspicious, and there's no particular reason to assume they're banned. There's also no particular reason to imagine there's been an arson attack. Players can perfectly well assume that this means "you don't find anything important".
  • You call for a roll. Result: this adds uncertainty in the wrong place, because the real question is whether they work out the connection. If they roll badly, they won't have the information needed to draw the conclusion, which they definitely ought to have. At the same time, you are flagging up that the matches are important clues.
  • You mention at the start of the game that matches are not allowed in school and there was a fire in the chemistry lab last term. Result: players keep a sharp eye out for matches, and for anything at all relating to the chemistry lab. As soon as they find the labcoat, they will assume it is important and draw the obvious conclusion.
  • When they find the matches, you also tell them: "Matches aren't allowed in school. You remember that there was a fire in the chemistry lab last term." Result: this is more than a clue, it's a chunk of plot that limits the players' involvement. If it's correct, they're basically not getting to do much investigating. If it's not, you're dropping massive red herrings.

If you give out only the relevant information at the start, you're highlighting the things they should look out for. If you give out a mixture of real info and red herrings, this might really annoy the players. If you give out a lot of general information and only a little is relevant, then for one thing it might be quite boring to wade through before the game starts, and for another, they might well ask at the end "wait, so after all that, the school pig was completely irrelevant?".

In a similar way, people become a problem. They will know most of the other people in the school, at least by sight. If twenty people pass you on the way to class, and the Keeper says "one of them is Jones", that's a fairly obvious hint that Jones is important. If the Keeper names all of them, that's boring. This makes it difficult to draw someone into the plot without making it obvious that they are a Named NPC.

Any solutions?

I'm not really sure what's the best way to tackle these problems, at least in a concrete way. It will depend on the scenario, and I suspect the timeframe is important. If you have a slow-burning plot, investigators can be gently led around the area and naturally encounter many of the locations. If you start in media res, things will be far more difficult.

My idea was to do an initial briefing with maps available, just so everyone's available of what exists within the school at all, and then to give more description when they were visited. It's going to be fairly obvious which areas get singled out and which are lumped into one shared description, but that's how it goes. I'd also suggest encouraging them to ask questions about the school and its history before the game starts, just to help fix things in people's minds, as well as to visit them during the early stages of the game, so they have a good idea of what the school is like.

People are a lot harder, because you just can't give people a useful sense of multiple characters the way you have in real life, without any actual interaction. The same applies to recent events, which are mostly irrelevant but might contain significant plot points. I'd suggest a similar tack, with an initial overview, and some handouts to help ease the pain. For other information, I'd try to call for Know and Idea rolls whenever it seems there might be a relevant link, and be fairly generous with those.

I'm not really sure what can be done about everyone being trapped. Basically, once you're under suspicion, there's not much you can do about things. This seems to call for a design-level solution, so that simply guessing who might be involved in a situation isn't enough to solve it, and won't put too much of a dent in the NPCs' plans.


  1. On a tangentially related not, I think that I've actually visited most tourist attractions in my hometown, it's just that I've never actually visited them *as* tourist attractions. I think one's interactions with local landmarks can broadly be categorised as:

    * Places you genuinely never go because you're not interested
    * Places you go so often that you forget that most people don't hang out in places like this all the time
    * Places you only go when something specific draws you there.

    I think the last two of these have interesting implications for scenario design, specifically for the difference between location-based and event-based scenarios. I think that a lot of these problems arise because it's hard to take a *location based* approach to a scenario set in a familiar location because, well, the location is already familiar. I think this becomes a particular problem with prewritten scenarios because it is very tempting to lavish a lot of time and attention on *detailing* the location, which would seem to support location-based gameplay.

    A lot of the core problems disappear in an event-based game. Of course pupils at St Niggurath's School for Wayward Youngsters have no reason to go exploring the Dark Spooky Woods that are an everyday part of their school experience, but if a fellow student goes missing in those woods, they have a reason to go and look for them.

    Shall comment further later.

    1. Ah, interesting. Yes, I think you’re right.

  2. Further comments:

    I don't think the problem of dissonance is as bad as it seems, actually. Yes it's true that if there's a fire in an RPG and then a mysterious event, players are quite likely to think the two are related in ways that real people might not, but I think that's no truer in a familiar setting than an unfamiliar one. If the players hear on the news that a building has just burned down in the town they're visiting, they will assume that every mysterious event has something to do with the burned building.

    I think, for me, there's not so much dissonance here as concordance. If an event is outlandish enough that my character would definitely have noticed it, then it isn't unreasonable that they might connect it with whatever is going on mystery-wise, so my IC familiarity with the school and my OOC ability to spot pertinent facts actually reinforce each other here.

    It becomes more difficult in the (I think more niche) situation in which something that would seem very outlandish to me is natural to my character (what, don't all boys' schools have nightly blood rituals on the tennis courts?) although at that point dramatic irony tends to kick in.

    I suspect part of it is a mileage thing. I think I have a much higher tolerance for metagaming than you do and so I don't mind being in a situation where I understand out-of-character the significance of something I do not understand in-character. To me there's no difference between knowing that, because the GM told us there was a fire last term, that the fire and who started it are probably significant than knowing that, because I am playing Call of Cthulhu, my character will probably encounter the Cthulhu mythos at some point.

    1. You’re probably right about metagame tolerance. However, I think you’re looking for dissonance in a slightly different place. What I’m talking about isn’t dissonance between the player and PC in making connections, but between their knowledge and understanding of a situation. There’s always a certain amount of this, but usually it’s pretty mild.

      “I don't mind being in a situation where I understand out-of-character the significance of something I do not understand in-character.” That’s not entirely what I intended to get at. Um, maybe it’s more about the manner than the fact, or something?

      I think one of the issues is not so much that players will pick up on things readily, but that it becomes difficult for information to be introduced in a way that isn’t heavy-handed and GM-driven, and this is unsatisfying.

      For example, if I visit Innsmouth and the town hall was recently burned down, there are various ways I could encounter this information as a natural outcome of my decisions. I might do some research on the town and find news articles about the fire. I might talk to someone in a pub and pick up gossip. I might just walk past the charred remains. I might find a paper in the hotel that mentions the fire, perhaps while discussing a topic that’s more obviously connected to my reason for visiting. In all these cases, the player and the character discover the information together; some are more player-driven than others, but they’re not particularly infodumpy either.

      If we start a school game in November and the chemistry lab burned down last month, then to my mind it’s difficult to introduce that information under the players’ control, because the characters already know it. Trying tactics like those above – the pupils walk past the charred classroom – feels unnatural to me, it’s all very “as you know, your father the burned-down chemistry lab”.

      It’s not so much a question of flagging up importance (anything the GM mentions is liable to be significant) as a jarring discrepancy between what my character knows and what I know. As I said, it follows the Sherlock Holmes template, where facts available to the detective are not available to the reader/player. It leaves you waiting for the key facts to be disbursed by the GM. Of course, you’re always ultimately dependent on the GM telling you things, but this kind of situation seems to make it much more obvious. It’s... inelegant, I suppose.

      Not sure if that makes any sense. It’s quite hard to articulate what I mean, sorry.

    2. I think I get where you're coming from, it's just that I don't see clues that come out in play to be particularly more organic than clues that come out in backstory.

      I'm generally perfectly happy with a quick pregame infodump that just goes "for what it's worth, there's a well-connected old boy who went missing last year, the chemistry lab burned down last month, and there's weird inscriptions on the portraits in the dining hall". I'd find it no more jarring than finding a newspaper in play that mentioned any one of those three things.

    3. Whereas my gut reaction to that is "great, so, what do you need me here for again?"

      Obviously in an actual game I'd be rather more tolerant, but I suspect I have much higher tolerance for both setting infodumps and metagame tropes than I do for metagame infodumps. It sort of feels like I'd be reduced to walking through the game looking for ways to tick the aforementioned plot boxes, and I'd rather pick the clues up in play and hide the gears a bit.

      Looks like just a case of YMMV.

    4. I think it might be quite a deep-seated difference in preferences actually, since I actually mildly favour metagame infodumps to IC infodumps, for roughly the same reason I prefer exposition in novels to be in narration rather than dialogue. I tend to feel that I'd rather have information given to me upfront, particularly if it's information I would, in character, have known from the beginning.

      I think it ties back into the conversation we were having a while back about whether the actual getting of information is interesting, or whether things get interesting once you have the information.

    5. Well, I'm not suggesting having IC infodumps at all, although that might just be a terminology thing. I'd certainly rather have IC knowledge OOC from the beginning, as I said above, otherwise things get weird.

      Broadly speaking, it seems like you'd prefer prior plotpoint knowledge to be given as an OOC infodump before the session rather than as an IC infodump during the session.

      My position is more that prior plotpoint knowledge is inherently inclined to be inelegant (I *think* I think that because it's so difficult to limit the metaness of it) and I'm not very convinced by either IC or OOC solutions I've seen so far. I'm wary about the infodump, be it IC or OOC, because for some reason metagame knowledge that certain facts are plot-relevant bothers me more than metagame knowledge that the reclusive noble Baron Alucard is probably a wrong'un. Maybe because playing along with tropes feels to me like it offers guidance, whereas knowing plotpoints feels like it's pushing me to do things without IC justification, in a kind of point-and-click way? I dunno, it's late.

      whether the actual getting of information is interesting [finding clues], or whether things get interesting once you have the information [putting clues together].

      It's kind of the inverse of this. I think I'm suspicious that if I've been given too much plot ahead of time, it will reduce my mileage in putting the pieces together, and just leave me with the job of getting the information. I get a lot of enjoyment out of just meandering through gameworlds poking the scenery and putting things together for myself.

      Coming back to your first post, I've no problem with an event-based game with an OOC briefing like: "You're at Warthogs, your friend Basil has gone missing and was last seen near the Dark Scary Woods, and yesterday he was being mysterious and carrying around an old book he was very protective of." That is a coherent set of plot knowledge that my character is aware of and a good starting point.

      I’d be less happy with something like: "You're at Warthogs, your friend Basil has gone missing and was last seen near the Dark Scary Woods, and yesterday he was being mysterious and carrying around an old book he was very protective of. Miss Williams has a dragon-slaying spear in her office. There’s mysterious writing around the mirror in the second-floor toilets. Gruzelbean, the famous Founder of Warthogs, was exceptionally fond of cabbage. A famous old boy went missing last year in Cardiff." Most of those aren’t obviously significant to the character. I feel like I’d be wandering through the game with an eye out for possible connections to any of those plot points, rather than concentrating on what’s in front of me.

      I suspect this may link in with our different takes on IC dialogue and that sort of thing. Maybe I’m just less meta.

    6. I’d be less happy with something like: "You're at Warthogs, your friend Basil has gone missing and was last seen near the Dark Scary Woods, and yesterday he was being mysterious and carrying around an old book he was very protective of. Miss Williams has a dragon-slaying spear in her office. There’s mysterious writing around the mirror in the second-floor toilets. Gruzelbean, the famous Founder of Warthogs, was exceptionally fond of cabbage. A famous old boy went missing last year in Cardiff."

      I think the difference there, though, is more in the kind of information that each point represents, because not all of the things on that list are things that you would necessarily expect your character to actually know. I mean I wouldn't want a plot dump like that either, because it would seem a bit random and restrictive.

      I *think* the problem with this kind of dump is that it basically says "these are the only things worth knowing" by dint of its being so exhaustive and list-like. I think I'd be inclined to rephrase it to something like:

      "Okay, so your characters will all be students at Warthogs school of SUBJECTS TOTALLY UNRELATED TO WITCHCRAFT AND/OR WIZARDRY. It's sort of a bit like Mallory Towers, only with a sort of kooky occult/lovecraftian vibe, so it's got all of the usual boarding school trappings, but there's bits of weirdness all over the place, like there's this mysterious writing around the mirror in the second floor toilets, and there's a rumour that one of the teachers has a dragon-slaying spear in her office.

      The place has a lot of weird traditions, like they serve cabbage at literally every meal, even breakfast, because Gruzelbean - the founder of the school - was really into cabbage.

      It's also the sort of place where nebulously Bad Things happen a lot. Students go missing, ex students go missing. There was a bit of a scandal last year because a particularly famous old boy went missing in Cardiff, and yesterday your friend Basil vanished near the Dark Scary Woods. You'd almost certainly all have noticed him carrying this old book around at the time, and being weirdly protective of it."

      It's basically the same information, but it doesn't make it so obvious that THIS IS ALL THE CLUES (and indeed it may not be).

      Incidentally, I think a good example of how to handle this kind of thing is the way Arthur handles Vana's Dwarflore in the D&D game. Basically he'll only mention things when they come up, but then he'll just say "you'll know that this is a dwarf thing/smith thing."

  3. Players will certainly also want to go to places that are well-detailed and seem interesting even if it's obviously just random fluff. Just like regular people can hear about a place and want to go there, so will PCs ... only they generally don't have the commitments we do which prevent random travels. This can also be an issue, when PCs rush off to a place (generally after an adventure) simply to see it.

    I also think that even if the PCs *know* the area, the players don't need to know it intimately. They can simply say: "Ooh, a bar matchbox. What do I know about this bar?" and take it from there. So the players can still uncover things as the characters do.

    As for me, I'd like a pre-game info dump as I'd imagine that knowing the lab burned down, a guy went missing, and there's markings on the portraits are just the tippy tip of an incredibly exciting iceberg. What am I here to do? Why, to figure out how they connect and to discover more strange mysteries. If this is what I glean upfront, imagine what I can find out later on?

    Alternatively, I could play along with rumours and myths coming up during the game that point out things that've happened. "I reckon the chem lab teacher burned it down because it was all old stuff and he wanted an upgrade...." says one student. "He snuck into the chem lab and burned it down. What do you think?" Then they turn to me to give my thoughts.